UNICEF Data Analysis Report

Introduction

Children’s rights and development are critical to building a nation’s future.
This report leverages UNICEF’s data to highlight key indicators across countries, focusing on:

Health outcomes
Educational progress
Nutritional status
Child protection measures

Our goal is to uncover meaningful trends that can guide effective, data-driven interventions.

Data Exploration Journey

To better understand how children’s well-being varies across the world,
we explore the available UNICEF indicators through a series of visualizations.

We start by examining which countries stand out based on their average observed values.
Then, we trace the evolution of these values over time, highlighting differences between males and females.
Finally, we zoom out to a global perspective with a colorful world map showing the latest data across nations.

These visual insights help reveal both progress and persistent challenges in child development across regions.

Code

# Import libraries
import plotly.express as px
import plotly.graph_objects as go
import pandas as pd
from plotnine import *
import matplotlib.pyplot as plt
import seaborn as sns

# Load dataset
df = pd.read_csv("unicef_indicator_2.csv")

# Clean: Group by country and take the mean (or max) observed value
df_clean = (
    df.groupby('country', as_index=False)
      .agg({'obs_value':'mean'})
)

# Sort and select top 10 countries
top10 = df_clean.sort_values(by='obs_value', ascending=False).head(10)

# For pie chart, calculate top 5
top5 = df.sort_values(by='obs_value', ascending=False).head(5)

# Display first few rows
df.head()

# Table of Top 10 Countries
top10 = top10.reset_index(drop=True)
top10.index = top10.index + 1
top10

	country	obs_value
1	Ethiopia	44.833333
2	Burkina Faso	41.500000
3	Chad	39.033333
4	Cameroon	38.900000
5	Togo	38.466667
6	Madagascar	36.700000
7	Haiti	35.233333
8	Niger	34.333333
9	Nigeria	31.500000
10	Burundi	30.933333

Focusing on the Leaders

To identify which countries have the highest average outcomes,
we visualize the Top 10 countries based on their observed values.
This helps highlight regions making significant progress in key child development indicators.

Summary of Average Observed Values by Country

The table below displays the average observed values across countries, providing a snapshot of key performers.
Identifying countries with the highest averages enables targeted focus for deeper analysis.

Interactive Visualization of Top 10 Countries

Code

fig = px.bar(
    top10,
    x='obs_value',
    y='country',
    orientation='h',
    text='obs_value',
    color='obs_value',
    color_continuous_scale='Blues',
    title="Top 10 Countries by Observed Value (Interactive)"  # Simple string here
)

fig.update_layout(
    title={
        'text': "Top 10 Countries by Observed Value",
        'x': 0.5,
        'xanchor': 'center',
        'font': {'size': 24}
    },
    margin=dict(t=100),
    xaxis_title='Observed Value',
    yaxis_title='Country',
    plot_bgcolor='white'
)

fig.update_traces(
    texttemplate='%{text:.1f}',
    textposition='outside'
)

fig.show()

Understanding Trends Over Time

Beyond averages, it’s crucial to explore how countries have evolved over time.
In the next visualization, we track the observed values across different periods, separated by gender,
to spot improvements, declines, or emerging patterns in children’s well-being.

Observed Values Over Time

To understand how progress unfolds over time, we plot observed values across different years, separated by sex.
This view highlights patterns, fluctuations, and trends that reveal the trajectory of child development indicators globally.

Code

# 1. Prepare Figure
fig = go.Figure()

# 2. Loop through each country and add Male and Female separately
for country in df['country'].unique():
    df_country = df[df['country'] == country]
    
    for sex, color in zip(['Male', 'Female'], ['blue', 'red']):  # Male in blue, Female in red
        df_sex = df_country[df_country['sex'] == sex]
        
        fig.add_trace(go.Scatter(
            x=df_sex['time_period'],
            y=df_sex['obs_value'],
            mode='lines+markers',         # <-- Keep lines + points
            name=f"{country} - {sex}",
            line_shape="spline",          # <-- Smooth curves
            line=dict(color=color, width=2), # <-- Make lines slightly thicker
            opacity=0.4                   # <-- Light opacity to reduce clutter
        ))

# 3. Update Layout
fig.update_layout(
    title={
        'text': "Observed Values Over Time: Male vs Female Across Countries",
        'x':0.5,
        'xanchor': 'center',
        'font': dict(size=24)
    },
    xaxis_title="Time Period",
    yaxis_title="Observed Value",
    template="plotly_white",
    hovermode="x unified",
    legend_title="Country - Sex",
    height=700,
    margin=dict(t=100, l=50, r=50, b=50),
)

# 4. Show figure
fig.show()

Deep Dive into Specific Countries

While general trends are useful, it’s also important to focus on particular countries individually.
Here, we zoom into selected countries to better understand their unique journeys.
A scatter plot with a linear regression line helps highlight the overall direction of change — whether progress is accelerating, slowing down, or remaining steady.

Trend of Observed Values by Country

Beyond static averages, it’s essential to explore temporal trends at the country level.
The scatter plot below includes a fitted regression line to summarize the overall trend across selected countries.

Code

# 📌 First: Filter your DataFrame for specific countries
selected_countries = ["Cameroon", "Ethiopia", "Haiti", "Nigeria"]
df_filtered = df[df['country'].isin(selected_countries)]

# 📌 Second: Create the scatter plot with a regression line
import plotly.express as px

fig = px.scatter(
    df_filtered,
    x="time_period",
    y="obs_value",
    color="country",
    trendline="ols",  # Add ordinary least squares regression
    trendline_scope="overall",  # Single regression line for all data
    title="Scatter Plot with Linear Regression Line",
    labels={
        "time_period": "Time Period",
        "obs_value": "Observed Value",
        "country": "Country"
    },
    template="plotly_white"
)

# 📌 Third: Beautify the chart layout
fig.update_layout(
    title={
        'text': "Scatter Plot with Linear Regression Line (Combined)",
        'x': 0.5,
        'xanchor': 'center',
        'font': dict(size=24)
    },
    height=600,
    legend_title="Country",
    margin=dict(t=80, l=50, r=50, b=50),
)

# 📌 Fourth: Show the final figure
fig.show()

Global Overview of Observed Values

After exploring individual country trends, it’s valuable to step back and look at the bigger picture.
The following world map presents the latest observed value for each available country.
This visualization helps identify regional disparities, highlight clusters of progress, and spot areas where targeted interventions might be most needed.

Code

# Take the latest value for each country (even if years differ)
df_latest = df.sort_values('time_period').groupby('country').tail(1)

fig = go.Figure(data=go.Choropleth(
    locations=df_latest['country'],
    locationmode='country names',
    z=df_latest['obs_value'],
    colorscale='Viridis',
    colorbar_title='Observed Value',
    marker_line_color='white',
    marker_line_width=0.5,
))

fig.update_layout(
    title_text='World Map of Observed Value (Latest Available per Country)',
    title_x=0.5,
    geo=dict(
        showframe=False,
        showcoastlines=True,
        coastlinecolor="white",
        showland=True,
        landcolor="rgb(217, 217, 217)",
        oceancolor="rgb(173, 216, 230)",
        showocean=True,
        projection_type='natural earth'
    ),
    template="plotly_white",
    margin=dict(t=50, l=0, r=0, b=0),
)

fig.show()

Animated Rotating Pie Chart of Top 10 Countries by Observed Value

This visualization presents an interactive and animated pie chart displaying the top 10 countries based on their observed values.
Built using Plotly, the chart starts with a static view and offers a “Play” button to trigger a smooth 360-degree rotation animation.
Each frame of the animation gradually rotates the pie chart, providing a dynamic and engaging way to explore the distribution of values among the countries.
The chart also features a clean design with a slight hole in the center (creating a donut chart effect) and a clear title to enhance readability.

Code

import numpy as np

labels = top10['country']
values = top10['obs_value']

fig = go.Figure()

# Add the initial pie chart
fig.add_trace(
    go.Pie(labels=labels, values=values, hole=0.3, rotation=0)
)

# Create 30 frames rotating the chart
fig.frames = [
    go.Frame(
        data=[go.Pie(labels=labels, values=values, hole=0.3, rotation=angle)],
        name=str(angle)
    )
    for angle in np.linspace(0, 360, 30)
]

# Add play button for animation
fig.update_layout(
    title_text="Top 10 Countries by Observed Value",
    title_font_size=24,
    updatemenus=[{
        "type": "buttons",
        "buttons": [{
            "label": "Play",
            "method": "animate",
            "args": [None, {"frame": {"duration": 100, "redraw": True}, "fromcurrent": True}]
        }]
    }],
    margin=dict(t=100, b=50),
)

fig.show()

Conclusion

This report provides an analytical overview of key child development indicators across countries, using UNICEF’s valuable dataset.
Through interactive visualizations, we identified both high-performing regions and areas needing urgent intervention.
Future efforts should focus on sustaining gains, closing regional gaps, and ensuring that no child is left behind.